Quick Critique of Gary North

I really don't want to devote too much time to this--I've already wasted an arbitrarily long amount of time on the sneezing riddle. But Gary North was such a jerk to some young gold bug, that I feel I must weigh in.

* North has been patting himself on the back ever since the February CPI growth was flat, and then again when gold prices fell in early April. He takes credit for these outcomes because--unlike most other libertarian economists--North has said that we are in a deflationary environment. He said this at least as far back as April 2007, where he explained that M1 is the best predictor of price inflation, and based on that, we should have low, possibly negative, price inflation.

* First, North's argument to show the superiority of M1 is just plain goofy. His method is to look at the total percentage growth in various monetary aggregates over different time frames, and them compare this with the percentage change in CPI. All of the aggregates overshoot, if memory serves, but M1 does the least. Hence M1 is the best predictor of inflation, and since M1 has been flat for a while, February's flat CPI was just what North was warning us about!

* That is goofy, as I said above. There are a lot more sophisticated ways you could use monetary aggregates to try and predict price movements. E.g. you could try correlating changes in an aggregate with changes in price. To simply look at gross percentage increases over a given time interval is very crude. To give you a ridiculous example, North's method would "prove" that in some settings, me telling you the Celsius temperature would be a poor predictor of the Fahrenheit temperature. (Suppose in year 1 it's 10C, i.e. 50F. Then the next year it warms to 20C, i.e. 68F. The Celsius reading doubled, while the Fahrenheit went up 36%. So the Celsius scale tends to overpredict changes in Fahrenheit; we need to find a better predictor of Fahrenheit scale changes!)

* More important, M1 has been flat before. Look at it here in the mid-1990s. Did we have deflation for a years back then? Of course not. So why should we expect it now?

* What is absolutely ABSURD about North's taking credit for the February CPI number is that he made his "maverick" prediction back in April 2007. I.e. North was calling for deflation four months into the calendar year with the highest inflation (4.1%) in over a decade.

* And if you want to talk about individual months, that's fine. Well after North's prediction, where he warned his readers to watch out for deflation, there was an annualized monthly increase in the CPI of over 10%. This was the 2nd most inflationary month going back at least 17 years. (From the graph I can't quite make out which month; it looks like it was either October or November of 2007.) I.e., if North wants to take credit for the 0% annualized CPI growth in February 2008, then he also has to explain the 10%+ annualized CPI growth in late 2007. Again, he called for deflation back in April 2007.

* Finally, North was tearing Seville a new one over the fall in gold on April 1. Well, do you think it possibly had to do with the change in quarter? North writes: "Of course, he can wait until gold rebounds above $1,000 and the CPI goes back to 4% a year." Well, gold's already risen about $10/oz. since North wrote that.

* In conclusion, I'm not saying I liked Seville's analysis. But I think North was WAY too cocky in his treatment of the young punk. I'm also not predicting that gold or inflation will go through the roof in 2008. (I think if the credit markets recover, then those things will indeed happen, since the Fed is primed to flood the system with new money, plus the recession will reduce the flow of real output.) But even if 2008 turns out to be a year of moderate inflation, I'm not convinced North should be credited with a brilliant prediction.

Comments

  1. Isn't the real problem here that North is ignoring the principle of our common mentor, Mises: There are no constant relationships in these matters? M1 might relate to inflation in one way for 20 years, and then the relationship will completely alter. As Tony Lawson, who is a macro lecturer at Cambridge, told me, "I've never seen a macro model that didn't break down as soon as it was promulgated."

    These studies can be useful for pattern prediction, as long as the predictor is aware that the pattern is a contingent product of historical circumstances, and not a "law of economics."

    ReplyDelete
  2. Anonymous8:12 AM

    North also describes himself as a "gold bug" and has stated he will never sell his gold stash in his lifetime. He is a typical investment newsletter writer; he will be right no matter where gold prices go.

    ReplyDelete

Post a Comment

Popular posts from this blog

Libertarians, My Libertarians!

"Machine Learning"

"Pre-Galilean" Foolishness